ZK Co-processor: Building the Bridge to Large-scale Web3 Applications

In the traditional computer field, coprocessors are responsible for handling other complex tasks for the CPU. Common coprocessors include Apple's M7 motion coprocessor and Nvidia's GPU. Coprocessors can offload complex code with high single-performance requirements, allowing the CPU to handle more flexible and varied parts.

There are two serious constraints on the development of applications on the Ethereum chain:

1. High gas fees limit the scope of on-chain application development.
2. Smart contracts can only access recent block data, making it difficult for data-driven innovative applications to emerge.

These two issues severely limit the emergence of "mass application" products. The ZK co-processor can serve as an auxiliary processing unit for Ethereum, handling compute and data-intensive tasks, thereby addressing these issues.

The application boundaries of ZK co-processors are very broad, covering almost all real dapp application scenarios, such as social networking, gaming, DeFi, risk control systems, data storage, and large model training. Theoretically, all the functionalities that Web2 applications can achieve can also be implemented on the blockchain with ZK co-processors, while leveraging Ethereum as a settlement layer to ensure application security.

Currently, well-known co-processor projects in the industry are mainly divided into three categories: on-chain data indexing, oracle, and ZKML. Among them, the general-purpose ZK co-processor (General-ZKM) can cover all these application scenarios.

General ZK Collaborative Processor Technology Architecture

Analyzing the technical architecture of general-purpose ZK co-processors using Risc Zero, Lagrange, and Succinct as examples:

Risc Zero

The ZK coprocessor of Risc Zero is called Bonsai, which is a set of blockchain-agnostic zero-knowledge proof components. Its main functions include:

1. General zkVM, capable of running any virtual machine in a zero-knowledge/verifiable environment.
2. ZK proof generation systems that can be directly integrated into smart contracts or chains.
3. General rollup can distribute the computations proven on Bonsai to the chain.

The components of Bonsai include the prover network, request pool, Rollup engine, mirror center, state storage, and proof market, etc.

Lagrange

Lagrange aims to build co-processors and verifiable databases, including blockchain historical data, to support the development of compute and data-intensive applications. Its main features are:

1. Verifiable Database: Indexed on-chain smart contracts storage, reconstructing blockchain storage, state, and blocks.
2. Computation based on the MapReduce principle: using data separation for multi-instance parallel computation.

Lagrange designed new contract storage, account states, and block data structures to support SNARK/STARK proofs. Its ZKMR virtual machine performs distributed computation and proof aggregation through two steps: Map and Reduce.

Succinct

The goal of Succinct Network is to integrate programmable facts into various parts of the blockchain development stack. Its feature is to prove market compatibility with various proof systems.

The off-chain ZKVM of Succinct is called SP( Succinct Processor ), supporting LLVM languages such as Rust. Its core features include:

1. Recursion proof technology based on STARKs
2. Wrapper from SNARKs to STARKs
3. Precompiled Centralized zkVM Architecture

Similarities and Differences Between Co-Processors and Layer 2

Unlike user-oriented Layer 2, co-processors are application-oriented and can be applied as acceleration components or modular components in:

1. Off-chain virtual machine of ZK Layer2
2. Off-chain computing power of public chain applications
3. Cross-chain Data Oracle
4. Cross-chain bridge messaging

The co-processor brings the potential for real-time synchronized data across the entire chain and high-performance, low-cost trusted computing, which can reconstruct most middleware of the blockchain.

Challenges Faced by Co-processors

1. High barrier to entry for developers
2. The track is in its early stages, and the landscape is unclear.
3. Prerequisites such as hardware have not been implemented.
4. The technical paths are similar, making it difficult to establish technical barriers.

Summary and Outlook

ZK co-processors are expected to reconstruct the development paradigm of Web3 applications, achieving a fully chain real-time provable database and low-cost off-chain computation. The implementation of ZK computing chips is a prerequisite for large-scale commercial applications. It is anticipated that the next cycle will see the commercial implementation of the ZK industry chain, truly supporting on-chain interactions at the level of 1 billion users.